INDEX

Explanations

words related to legal matters and government statements

oai_token-act-pair · gpt-3.5-turbo

punctuation and formatting cues

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 spir

-1.00

 spitting

-0.91

 sculpt

-0.90

 experimenting

-0.89

 cooking

-0.89

 trendy

-0.88

 branching

-0.86

 flattened

-0.85

 swall

-0.85

 sucker

-0.85

POSITIVE LOGITS

Statement

1.76

Introduction

1.72

Dear

1.72

Executive

1.61

Section

1.60

References

1.59

CHAPTER

1.58

Acknowled

1.58

Subject

1.58

Thank

1.56

Activations Density 0.296%

words related to legal matters and government statements

punctuation and formatting cues

No Comments

No Known Activations